ARTICLE

pubs.acs.org/EF

CO2 Capture over K2CO3/MgO/Al2O3 Dry Sorbent in a Fluidized Bed

Lei Li,†,‡ Yong Li,†,‡ Xia Wen,† Feng Wang,† Ning Zhao,† Fukui Xiao,† Wei Wei,*,† and Yuhan Sun*,†,§
†
State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan 030001,
People’s Republic of China
‡
Graduate University of the Chinese Academy of Sciences, Beijing 100049, People’s Republic of China
§
Low Carbon Conversion Center, Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201203,
People’s Republic of China

ABSTRACT: CO2 capture was carried out over dry K2CO3/MgO/Al2O3 sorbent in a fluidized-bed reactor. The result showed that
the total CO2 adsorption capacity of the new sorbent reached 109.6 mg of CO2/g of sorbent in the bubbling regime at 60 °C. The
X-ray diffraction (XRD) results revealed that there were several kinds of carbonates formed during CO2 adsorption, such as KHCO3,
KAl(CO3)2(OH)2, K2Mg(CO3)2, K2Mg(CO3)2 3 4H2O, and MgCO3. The high concentration CO2 could be obtained by the multi-
stage adsorption/desorption cycles. The sorbent could be completely regenerated at 480 °C, and the adsorption capacity remained
unchanged after 6 cycles. Furthermore, the application of a conceptual CO2-capture process using this sorbent was proposed for an
existing coal-fired power plant.

1. INTRODUCTION However, the CO2 adsorption capacity (59.8 mg of CO2/g of

Carbon capture and storage (CCS) is recognized as a promis-
ing option for CO2 emission reduction.1 6 The postcombustion
process is more suitable for CO2 capture for the traditional
pulverized coal power plants.6 9 Until now, several processes
have been studied for the postcombustion carbon-capture pro-
cess, such as absorption with a solvent, adsorption using molec-
ular, membrane separation, and cryogenic fractionation.
However, these methods need to overcome the limits of cost
and consume a lot of energy.10,11 As a result, CO2 chemical
sorption using dry regenerable alkali-metal-based solid sorbents
has been proposed as an innovative concept for CO2 removal.
Alkali metal carbonates (K2CO3 or Na2CO3) could be used as a
sorbent, which reacts with CO2 and H2O to form alkali hydrogen
carbonates.7,10 25 Researchers have also tried to develop the
K-based sorbents supported on various supports, such as acti-
vated carbon, TiO2, MgO, ZrO2, Al2O3, CaO, SiO2, and zeolites.
The results of CO2 sorption capacities over different K-based
sorbents is summerized in Table 1.11 13,16,21 25 Among various
supports, MgO-supported K2CO3 could adsorb a large amount
of CO2 at low temperatures.11,21 However, the experimental
conditions were different from industrial applications (8 17 vol
% steam and 10 15 vol % CO2 in flue gases of a coal-fired power
plant), and the carbonation reaction rate was rather slow.26,27
Recently, Zhao et al. reported that K2CO3 calcined from KHCO3
with a hexagonal structure showed a fast rate of carbonation
reaction.26 28 Nevertheless, the mechanical strength of the
K2CO3/Al2O3 sorbent (314 N/cm2) was not high, which limits
its industrial application. The CO2-capture capacities were also
overestimated because they used thermogravimetric analysis
(TGA) for CO2 analysis, which cannot differentiate between
CO2 and H2O.
In our previous paper, it was found that the MgO/Al2O3
sorbent at 10 wt % MgO loading showed better performance for
CO2 capture at low temperatures in the fixed-bed reactor.29
sorbent) needs to be further improved. Furthermore, in the fixed-
bed reactor, heat control is difficult because the adsorption
process is highly exothermic. High superficial velocity is necessary
to reduce the reactor size. To solve these problems, a fluidized bed
is proposed as a proper process to control high volumes of flue
gases.9 13 Thus, in the present work, a new sorbent, MgO/Al2O3-
supported K2CO3, was prepared. The physical properties of the
sorbent, such as pore size, pore volume, and surface area, were
measured by nitrogen adsorption, and mechanisms of the adsorp-
tion and desorption were investigated with the aid of powder X-ray
diffraction (XRD) and temperature-programmed desorption (TPD).
The performance for CO2 capture was investigated in the
fluidized bed. The parameters, such as pressure drop, gas velocity,
and adsorption temperature, were also studied.
2. EXPERIMENTAL SECTION
2.1. Material Preparation. The sorbent used in this study was
prepared by the impregnation method. The preparation method of the
MgO/Al2O3 support (10 wt % MgO) and its microscopic structure were
reported in a previous paper.29 The new sorbent was prepared by
impregnating K2CO3 on MgO/Al2O3 using a wet muti-step impregnat-
ing method. Reagent-grade KHCO3 was supplied by Tianjin Fuchen
Chemical Reagents Factory. The preparation procedure of the sorbent
consisted of three steps: mixing and impregnation, drying at 110 °C for
12 h, and then calcining the solid product in a furnace under N2 flow
(20 mL/min) at 400 °C for 4 h. The sorbent was denoted as
KMgAlI3010, where K represents K2CO3, I represents the impregnation
method, and 30 and 10 represent the loading of K2CO3 and MgO,
respectively. In comparison, the sorbent without MgO was prepared by
the same method and was denoted as KMgAlI3000.
Received: April 2, 2011
Revised: July 4, 2011
Published: July 06, 2011

r 2011 American Chemical Society 3835 dx.doi.org/10.1021/ef200499b | Energy Fuels 2011, 25, 3835–3842
Energy & Fuels ARTICLE

Table 1. CO2 Sorption Capacities over K-Based Sorbents from the Literature
total capacity regenerator
temperature (mg of CO2/g temperature
material (°C) conditions of sorbent) methoda adsorber (°C) reference

K2CO3/ACb 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 86.0 TCD fixed bed 150 11
K2CO3/Al2O3 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 85.0 TCD fixed bed 350 11
K2CO3/USY 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 18.9 TCD fixed bed 11
K2CO3/CsNaX 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 59.4 TCD fixed bed 11
K2CO3/SiO2 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 10.3 TCD fixed bed 11
K2CO3/MgO 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 119.0 TCD fixed bed 400 11
K2CO3/CaO 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 49.0 TCD fixed bed 11
K2CO3/TiO2 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 83.0 TCD fixed bed 150 11
K2CO3/Al2O3 60 CO2, 1 vol %; H2O, 9 vol % 86.3 TCD fixed bed above 300 22
Re-KAl(I)30c 60 CO2, 1 vol %; H2O, 9 vol % 82.0 TCD fixed bed below 200 22
K2CO3/TiO2 60 CO2, 1 vol %; H2O, 9 vol %; 40 mL/min 90.0 TCD fixed bed 200 23
K2CO3/MgO 60 CO2, 1 vol %; H2O, 11 vol % 131.3 TCD fixed bed 350 400 21
K2CO3/ZrO2 50 CO2, 1 vol %; H2O, 9 vol % 91.6 TCD fixed bed 130 200 24
K2CO3/Al2O3 60 CO2, 13 vol %; H2O, 13 vol %; 2.3 m3/h TGA fluidized bed 350 12
sorbKX35d 70 CO2, 14.4 vol %; H2O, 7 vol %; 60 mL/min 90.0 TGA fluidized bed 150 230 13
sorbKX35d 60 CO2, 10 vol %; H2O, 12.2 vol %; 2900 mL/min 111.4 gas analyzer fluidized bed 300 16,25
a
TCD, thermal conductivity detector; TGA, thermogravimetric analysis. b Activated carbon. c Modified K2CO3/Al2O3 sorbent. d Commercial sorbent
(K2CO3, 35 wt %; supporters, 65 wt %).

2.2. Material Characterization. TPD of CO2 was performed in a
U-shape quartz tube. Typically, 0.1 g of KMgAlI3010 sorbent saturated
with simulated flue gases was swept with flowing argon for 1 h at 50 °C to
remove the weakly adsorbed CO2. Then, the temperature was increased
from 50 to 600 °C with a heating rate of 10 °C/min under argon flow at
50 mL/min. The CO2 concentration was monitored by an AMETEK
mass spectrometer.
The elemental analysis was performed using the inductively coupled
plasma (ICP) optical emission spectrometer, Perkin-Elmer 3000 equipment.
The structural change of sorbents before/after adsorption was
examined by XRD. The XRD analysis was performed on a Bruker D8
Advance X-ray diffractometer with a Cu KR source at 40 kV and 40 mA.
The surface area of the sorbent was determined by nitrogen adsorption
with the Brunauer Emmett Teller (BET) method on a Micromeritics
ASAP-2020 instrument. The pore volume and pore size distribution
were obtained with the Bopp Jancso Heinzinger method from the
adsorption isotherm.
The crushing strength of the sorbent was measured by a digital
crushing strength tester (Jiangyan Global Instrument Factory KC-3C),
which was a dynamometer measuring the force progressively applied to
the solid object during the advancement of a piston. The applied force
increases until the solid breaks and collapses into small pieces and
eventually powder. The corresponding value of the collapsing force is
defined as the crush strength.29 The detailed properties of KMgAlI3010
and KMgAlI3000 sorbents are shown in Table 2.
2.3. Apparatus and Operation Description. Figure 1 shows
the flowchart of the fluidized-bed experimental system, which mainly
consists of three subsystems: the gas injection system, fluidized-bed
adsorption system, and CO2 analysis system. The fluidized-bed riser is a
0.650 m tall stainless-steel pipe with an inner diameter of 0.045 m, and
the lower mixing zone is a 1.210 m tall stainless-steel pipe with an inner
diameter of 0.108 m. Temperatures along the riser are measured at six
different points by E-type thermocouples at the height of 300, 630,
760, 920, 1250, and 1580 mm from the distributor plate. The
differential pressure in the riser is measured by the pressure sensor
(Testo, Germany).
Table 2. Texture Properties of KMgAlI3010 and KMgA-
lI3000 Sorbents
property KMgAlI3010 KMgAlI3000
2
BET surface area (m /g) a
51.89 75.24
pore space (cm3/g)a 0.27 0.37
average pore diameter (nm)a 21.05 19.55
crushing strength (N/cm2)b 998.0 963.0
K2CO3/KMgAlI3010 (wt %)c 25.43 27.95
a
Measured by a nitrogen adsorption BET method. b Measured by a
digital crushing strength tester. c Measured by an ICP optical emission
spectrometer.
Dependent upon fuel composition, operating conditions, and the
type of coal burned, the composition of flue gases in coal-fired power
plants is expected to be approximately 10 15 vol % CO2,1,30 32 8
17 vol % H2O,17,18,33 4 5 vol % O2, N2, and trace quantities of other gas
species (SOx and NOx).32 A simulated flue gas composition of 10 vol %
CO2, 12 vol % H2O, and 78 vol % air was used in this study. The steam
was introduced by passing the air through a temperature-controlled gas
bubbler containing water, and the feed pipeline was heated to avoid H2O
condensation. The H2O concentration was analyzed by a relative
humidity analyzer (Vaisala, Finland). For steam pretreatment, only air
passed through the gas bubbler, and as a result, the sorbent adsorbed a
certain amount of H2O before adsorption of CO2. In the adsorption
step, the simulated flue gases entered the fluidized-bed lower mixing
zone from a perforated distributor and CO2 in the flue gases was
captured by solid sorbents in a riser where sorbent particles and flue
gases were well-mixed with concurrent flow. The flow was monitored by
the rotameters located at the inlet of the column. The outlet stream first
flowed through a condenser to remove H2O, then passed through a filter
to remove the dust and, finally, passed through a gas analyzer (Vaisala,
Finland) to analyze the CO2 concentration online.
Desorption and purge processes were conducted in a fixed-bed heat
regenerator (16 mm in inner diameter and 250 mm in length). A small

3836 dx.doi.org/10.1021/ef200499b |Energy Fuels 2011, 25, 3835–3842

Energy & Fuels ARTICLE

Figure 1. Flowchart of the fluidized-bed experimental apparatus.

Table 3. Experimental Conditions and Resultsa

total capacity breakthrough breakthrough capacity
run temperature (°C) flow (m3/h) total time (min) (mg of CO2/g of sorbent) time (min) (mg of CO2/g of sorbent)

1 60 12 10 109.6 2.83 78.3

2 60 22 10 84.5 1.67 60.3
3 60 32 10 59.8 1.00 52.4
4 75 12 10 101.2 2.67 66.0
5 90 12 10 98.1 2.58 65.6
6 60 12 10 77.0 1.83 43.5
a
Runs 1 5, KMgAlI3010 sorbent; run 6, KMgAlI3000 sorbent; temperature, adsorption temperature; total time, absorption time to reach total
capacity; breakthrough time, absorption time to reach breakthrough capacity; feed gas, 10 vol % CO2, 12 vol % H2O, and 78 vol % air; and mass of the
sorbent, 1000 g.

amount of sorbent (7.2 g) saturated with CO2 in the fluidized bed was
packed into the regenerator. In the desorption process, the feed gas was
used as a carrier gas. High-concentration CO2 effluent was obtained
because of the desorption of CO2 at elevated temperatures. The CO2
concentration was monitored by a gas analyzer (Vaisala, Finland),
while a large amount of the KMgAlI3010 sorbent used for adsorp-
tion/desorption cycles in the fluidized bed was regenerated in a
muffle furnace.
The range of spherical particle sizes for the KMgAlI3010 sorbent is
1.0 1.8 mm, and the average particle size is 1.4 mm. For the sorbent
used in this study, the minimum fluidization flow rate is 11.2 N m3 h 1,
which was determined from the plot of the pressure drop. The minimum
value of the calculated flow rate from various correlation transitions to
the turbulent regime is 35.0 N m3 h 1;34 therefore, the simulated flue
gases passed through the riser at a flow rate between 12 and 32 N m3 h 1,
which is higher than the minimum fluidization flow rate and lower than
the transition flow rate to the turbulent regime. The total and break-
through capacities were determined by the same methods as in a
previous paper.29In this work, it took a long time (t = 10 min) to
saturate the sorbent, and the breakthrough capacity was calculated when
the effluent CO2 concentration reached 5% of its feed concentration.
The breakthrough curves showed a gradual increase of the Cout/Cin ratio
from 0 to 1 (Cout and Cin refer to the CO2 concentrations in the outlet
and inlet flue gases, respectively). The CO2 sorption capacity was
calculated from the CO2 sorption profile by considering the CO2 flow
3837
profile without any sorbent mounted in the fluidized bed (used as a
blank). All of the experimental conditions and results are listed in
Table 3, and the run numbers in the following text refer to those listed
in the table.
3. RESULTS AND DISCUSSION
3.1. Identification of the Structure of the KMgAlI3010
Sorbent. Figure 2 illustrates the CO2 TPD profiles of the
KMgAlI3010 sorbent. The TPD profile of the KMgAlI3010
sorbent showed five CO2 desorption peaks, indicating that five
kinds of substances were formed after CO2 adsorption.
Figure 3 shows the XRD patterns of fresh sorbent, sorbent
pretreated with H2O, and sorbent after CO2 adsorption. As can
be seen, only K2CO3 (PDF 27-1348) and Al2O3 (PDF 29-0063)
phases were found for the fresh sample (Figure 3a). It might be
because MgO was highly dispersed on the internal surfaces of the
mesopores in the samples, which is accordance with the previous
study.29 After pretreatment with H2O for 1 h, K4H2-
(CO3)3 3 1.5H2O (PDF 20-0886) was the most abundant phase,
while the peak intensity of K2CO3 3 1.5H2O (PDF 11-0655) was
quite weak (Figure 3b). The K4H2(CO3)3 3 1.5H2O phase was
considered as the main active species for capturing CO2.26 28
The XRD pattern of the sorbent after CO2 adsorption at 60 °C
dx.doi.org/10.1021/ef200499b |Energy Fuels 2011, 25, 3835–3842
Energy & Fuels
Figure 2. CO2 TPD profiles of the KMgAlI3010 sorbent.
Figure 3. XRD patterns of the KMgAlI3010 sorbent (a) before, (b)
after the pretreatment with 12 vol % of steam for 1 h, and (c) after
adsorption with simulated flue gases at 60 °C: +, K2CO3; 4, Al2O3; 1,
K4H2(CO3)3 3 1.5H2O; 0, K2CO3 3 1.5H2O; 3, KHCO3; [, KAl-
(CO3)2(OH)2; b, K2Mg(CO3)2; f, K2Mg(CO3)2 3 4H2O; and #,
MgCO3.
showed five phases (Figure 3c): KHCO3 (PDF 12-0292), KAl-
(CO3)2(OH)2 (PDF 21-0979), K2Mg(CO3)2 3 4H2O (PDF 29-
1017), K2Mg(CO3)2 (PDF 33-1495), and MgCO3 (PDF 08-0479).
3.2. CO2 Adsorption in the Fluidized Bed. Figure 4 shows
the bed temperature profiles at different heights of the fluidized
bed. The bed was preheated by hot air for 40 min to reach the
desired temperature. After the sorbent (1000 g) was introduced
into the fluidized bed, the bed temperature decreased first, then
increased, and reached the pretreatment temperature (around
60 °C). The axial temperature gradient of the bed was very small,
which indicated that the fluidized bed had excellent heat-transfer
efficiency. Then, the gas flow was switched to air containing
12 vol % H2O and maintained at 60 °C for 40 min to activate the
sorbent. After that, the gas flow was switched to the simulated
flue gases. Some variations in the temperature across the fluidized
3838
ARTICLE
Figure 4. Bed temperature distribution at different heights of the
fluidized bed.
Figure 5. Axial differential pressure profile of the fluidized bed during
the sorption process. Flow rate, 12 N m3 h 1; amount of sorbent, 1000 g.
bed were observed because of the exothermic nature of the
adsorption.35,36
The pressure drop profiles at three different positions in the
fluidized bed were also monitored during the sorption process, as
shown in Figure 5. The differential pressures at three positions
were about 570, 560, and 250 Pa, respectively, from the bottom
to the top section. These results might provide valuable data to
estimate the design of fan power.13
To investigate the effect of the flow rate of flue gases, the
breakthrough curves of CO2 were measured at a flow rate
between 12 and 32 N m3 h 1. Figure 6 illustrates the CO2
concentrations in the outlet stream against the adsorption time
under different flow rates. Both total and breakthrough capacities
of CO2 adsorption are listed in Table 3 (runs 1 3). It can be seen
that CO2 was completely removed from the flue gases in the early
stage (1 2 min) because no CO2 was detected in the outlet
stream. Then, the CO2 concentration in the outlet stream
gradually increased to the feeding level in the next 2 5 min.
While, in a blank experiment, the Cout/Cin ratio quickly recov-
ered to 1 when there is no sorbent placed in the fluidized bed. As
dx.doi.org/10.1021/ef200499b |Energy Fuels 2011, 25, 3835–3842
Energy & Fuels
Figure 6. Breakthrough curves of CO2 against the adsorption time at
various flow rates. Adsorption temperature, 60 °C; feed gas, 10 vol %
CO2, 12 vol % H2O, and 78 vol % air; and amount of sorbent, 1000 g.
Figure 7. Sorbent regenerated under nitrogen at various temperatures,
such as (d) 180 °C, (e) 300 °C, (f) 400 °C, and (g) 480 °C: +, K2CO3;
4, Al2O3; 3, KHCO3; [, KAl(CO3)2(OH)2; b, K2Mg(CO3)2; f,
K2Mg(CO3)2 3 4H2O; and #, MgCO3.
seen from Figure 6, a shift of breakthrough curves to a shorter
time was observed at a higher flow rate, which indicated that the
CO2 adsorption capacity of the sorbent was saturated faster as
the space time of the feed gases increased.13 In comparison,
KMgAlI3010 captured more CO2 at the flow rate of 12 N m3
h 1 than at a higher flow rate. CO2 captured was 109.6 mg of
CO2/g of KMgAlI3010 sorbent, which was calculated by
integrating the area between the adsorption curve and the
blank curve. However, the CO2 capacity of the KMgAlI3000
sorbent was 77.0 mg of CO2/g of sorbent at the same experi-
mental conditions (run 6 in Table 3). From these results, it could
be concluded that MgO played an important role in the high
CO2-capture capacity of the KMgAlI3010 sorbent. It might be
because MgO not only participated directly in CO2 adsorption
but also could help K2CO3 transform to K2Mg(CO3)2 3 4H2O and
K2Mg(CO3)2.11,21
3839
ARTICLE
Figure 8. CO2 concentration change during the desorption and purge
cycle in the fixed-bed heat regenerator. Adsorption temperature, 60 °C;
pressure, 101 325 Pa; flow rate, 80 mL/min; balance gas, N2; and
amount of KMgAlI3010 sorbent, 7.2 g.
The effect of the adsorption temperature on the capacity of
CO2 adsorption was also investigated. Both total and break-
through adsorption capacities of CO2 are listed in runs 1, 4, and 5
of Table 3. Clearly, the total adsorption capacities of CO2 were
109.6, 101.2, and 98.1 mg of CO2/g of sorbent at 60, 75, and
90 °C, respectively. The adsorption capacity of CO2 decreased as
the adsorption temperature increased because both physical and
chemical adsorption (usually) were exothermic.35,36 Even though a
lower temperature is favorable for the carbonation reaction, the
process should be operated above the dew point of H2O to
prevent water condensation.
To identify the properity of the substances formed after CO2
adsorption, the XRD patterns of the sorbents were collected after
regeneration at various temperatures, e.g., 180, 300, 400, and
480 °C (Figure 7). As the temperature increased to 180 °C, the
phase of KHCO3 disappeared (Figure 7d).11,15,21,37 41 The
diffraction peaks of KAl(CO3)2(OH)2 gradually decreased with
the temperature increase and disappeared completely at 300 °C,
which indicated that KAl(CO3)2(OH)2 was completely con-
verted to K2CO3 at 300 °C (Figure 7e),11,15,22,37 while K2Mg-
(CO3)2 3 4H2O and K2Mg(CO3)2 were found to completely
decompose to K2CO3 at temperatures of 300 400 °C (Figure 7f).
After regeneration at 480 °C in nitrogen, MgCO3 was completely
converted to MgO (Figure 7g). It was consistent with the
structure of the MgO/Al2O3 sorbent after adsorption of CO2,
which was ascribed to unidentate carbonate species.29 However,
no diffraction peaks of MgO were observed, which might be
because of the low MgO loading. The XRD profiles in Figures 3a
and 7g were almost the same, which indicated that the sorbent
could be completely regenerated at 480 °C. These results were
well in agreement with the TPD result.
3.3. Regeneration of the Sorbent. Besides the CO2-capture
capacity, the regeneration property is also one of the most
important factors that affect the application of the sorbent. A
total of 7.2 g of the saturated KMgAlI3010 sorbents (exposed to
CO2 in the flue gases at 60 °C, with a flow rate of 12 N m3 h 1)
was regenerated in the fixed-bed heat regenerator. Although pure
CO2 could be used for desorption in the actual situation, a
relatively high concentration of CO2 also could be obtained using
dx.doi.org/10.1021/ef200499b |Energy Fuels 2011, 25, 3835–3842
Energy & Fuels
the feed gases as a carrier. Figure 8 shows the CO2 concentration
change in the desorption and purge processes. The test was
carried out by measuring the concentration of CO2 desorbed
when the temperature ramping rate was 20 °C/min from 60 to
480 °C. There were two main desorption peaks of CO2. One was
attributed to KHCO3, and the other one, which was a composite
peak, could be ascribed to other carbonates. This result was
inconsistent with the CO2 TPD results, which might have
originated from the different CO2 analyzer used in the regenera-
tion cycle. As shown in Figure 8, the average concentration of
CO2 desorbed in the desorption process increased from 10 to 25
vol %, which indicated that a higher concentration of CO2 could
Figure 9. Total and breakthrough capacities of CO2 over the KMgA-
lI3010 sorbent for multiple adsorption/desorption cycles in the fluidized
bed. Flow rate, 12 N m3 h 1; feed gas, 10 vol % CO2, 12 vol % H2O, and
78 vol % air; balance gas, air; regeneration temperature, 480 °C; and
amount of KMgAlI3010 sorbent, 1000 g.
Figure 10. Block flow diagram of the conceptual KMgAlI3010 sorbent CO2-capture process in coal-fired power plants.
3840
ARTICLE
be obtained by the multi-stage adsorption/desorption cycles. In
the purge process, the sorbent was purged in argon at 480 °C for
60 min, during which the KMgAlI3010 sorbent could be
completely regenerated. However, in practice, any CO2-capture
process might not need complete regeneration. As seen from the
CO2 TPD results (Figure 2), about 80% of CO2 was released
below 350 °C, so that the regeneration temperature of 350 °C
might be sufficient for sorbent recycle.
For commercial application, the sorbent should not only
possess high capacity and high selectivity but also exhibit stable
performance over multiple adsorption/desorption cycles.29 In
this study, the sorbent used for adsorption/desorption cycles in
the fluidized bed was regenerated under N2 flow (20 mL/min) at
480 °C for 3 h in a muffle furnace. Figure 9 shows the total and
breakthrough capacities of CO2 over the KMgAlI3010 sorbent
for 6 adsorption/desorption cycles in the fluidized bed. The
stable adsorption/desorption performance suggests the sorbent
is promising for practical applications.
3.4. Process Considerations. A block flow diagram of the
conceptual KMgAlI3010 sorbent CO2-capture process in a coal-
fired power plants is thus proposed in Figure 10. After the
electrostatic precipitator, the hot flue gases (350 400 °C) enter
a denitrification unit. Through the heat exchanger, the tempera-
ture of the flue gases decrease to 130 160 °C. Then, the gases
pass through a desulfurization unit. After the desulfurization unit,
the temperature of the flue gases decreases to 50 100 °C.
Desulfurized flue gases are then introduced into the carbonator,
where CO2 reacts with the sorbent and forms carbonates. The
carbonator, regenerator, and activator system are visualized as
fluidized beds. The cooling system located at the bottom of the
carbonator can remove the adsorption heat to ensure the stable
operation of the system. Flue gases with lower CO2 concentra-
tions are separated from partially reacted sorbent and then
discharged. A portion of the partially reacted sorbent is recycled,
and the other parts are transferred to the regenerator. At the
regeneration stage, the steam or pure CO2 can be used as carrier
dx.doi.org/10.1021/ef200499b |Energy Fuels 2011, 25, 3835–3842
Energy & Fuels
gases. The hot flue gases from different sections of coal-fired
power plants or a part of superheated steam could be used as the
source of heat to regenerate the sorbent. For example, the
KMgAlI3010 sorbent can be regenerated by the hot flue gases
(400 600 °C) from an economizer.5,41 Regenerated sorbent is
mixed with the fresh sorbent in the activator system. The sorbent
is pretreated with steam to form active species and then transfers
back to the carbonator. The pure CO2 product is ready for other
applications or sequestration.
4. CONCLUSION
Experiments on CO2 capture in the fluidized bed showed that
the KMgAlI3010 sorbent had a high CO2 uptake of 109.6 mg of
CO2/g of sorbent at the flow rate of 12 N m3 h 1 at 60 °C under
simulated flue gas conditions, which was larger than those of
most K-based sorbents reported before at low temperatures.
MgO played an important role in the high CO2-capture capacity
of the KMgAlI3010 sorbent. During the CO2 adsorption process,
several kinds of carbonates formed, including KHCO3, KAl-
(CO3)2(OH)2, K2Mg(CO3)2, K2Mg(CO3)2 3 4H2O, and MgCO3.
High concentrations of CO2 could be obtained by the multi-
stage adsorption/desorption cycles, during which the KMgA-
lI3010 sorbent could be completely regenerated at 480 °C. In
addition, the sorbent had high mechanical strength and exhibited
good stability in 6 cycles.
’ AUTHOR INFORMATION
Corresponding Author
*Telephone: +86-351-4049612. Fax: +86-351-4041153. E-mail:
weiwei@sxicc.ac.cn (W.W.); yhsun@sxicc.ac.cn (Y.S.).
’ ACKNOWLEDGMENT
This work was supported by the Lu’an Group, the Knowledge
Innovation Programme of the Chinese Academy of Science
(KGCX2-YW-323), the Ministry of Science and Technology of
the People’s Republic of China (2009BWZ003), and the Na-
tional Natural Science Foundation of China (50976116).
’ REFERENCES
(1) Mikkelsen, M.; Jørgensen, M.; Krebs, F. C. The teraton chal-
lenge. A review of fixation and transformation of carbon dioxide. Energy
Environ. Sci. 2010, 3 (1), 43–81.
(2) Soares, J. L.; Oberziner, A. L. B.; Jose, H. J.; Rodrigues, A. E.;
Moreira, R. F. P. M. Carbon dioxide adsorption in Brazilian coals. Energy
Fuels 2007, 21 (1), 209–215.
(3) Belmabkhout, Y.; Sayari, A. Isothermal versus non-isothermal
adsorption desorption cycling of triamine-grafted pore-expanded
MCM-41 mesoporous silica for CO2 capture from flue gas. Energy Fuels
2010, 24, 5273–5280.
(4) Drage, T. C.; Blackman, J. M.; Pevida, C.; Snape, C. E. Evaluation
of activated carbon adsorbents for CO2 capture in gasification. Energy
Fuels 2009, 23, 2790–2796.
(5) Xiong, R. T.; Ida, J.; Lin, Y. S. Kinetics of carbon dioxide sorption
on potassium-doped lithium zirconate. Chem. Eng. Sci. 2003, 58 (29),
4377–4385.
(6) Thiruvenkatachari, R.; Su, S.; An, H.; Yu, X. X. Post combustion
CO2 capture by carbon fibre monolithic adsorbents. Prog. Energy
Combust. Sci. 2009, 35 (5), 438–455.
(7) Liang, Y.; Harrison, D. P.; Gupta, R. P.; Green, D. A.; McMichael,
W. J. Carbon dioxide capture using dry sodium-based sorbents. Energy
Fuels 2004, 18 (2), 569–575.
3841
ARTICLE
(8) Qi, G. G.; Wang, Y. B.; Estevez, L.; Duan, X. N.; Anako, N.; Park,
A. H. A.; Li, W.; Jones, C. W.; Giannelis, E. P. High efficiency
nanocomposite sorbents for CO2 capture based on amine-functional-
ized mesoporous capsules. Energy Environ. Sci. 2011, 4 (2), 444–452.
(9) Romeo, L. M.; Abanades, J. C.; Escosa, J. M.; Pano, J.; Gimenez,
A.; Sanchez-Biezma, A.; Ballesteros, J. C. Oxyfuel carbonation/calcina-
tion cycle for low cost CO2 capture in existing power plants. Energy
Convers. Manage. 2008, 49 (10), 2809–2814.
(10) Park, Y. C.; Jo, S. H.; Park, K. W.; Park, Y. S.; Yi, C. K. Effect of
bed height on the carbon dioxide capture by carbonation/regeneration
cyclic operations using dry potassium-based sorbents. Korean J. Chem.
Eng. 2009, 26 (3), 874–878.
(11) Lee, S. C.; Choi, B. Y.; Lee, T. J.; Ryu, C. K.; Ahn, Y. S.; Kim,
J. C. CO2 absorption and regeneration of alkali metal-based solid
sorbents. Catal. Today 2006, 111 (3 4), 385–390.
(12) Zhao, C. W; Chen, X. P.; Zhao, C. S. Multiple-cycles behavior of
K2CO3/Al2O3 for CO2 capture in a fluidized-bed reactor. Energy Fuels
2010, 24, 1009–1012.
(13) Yi, C. K.; Jo, S. H.; Seo, Y.; Lee, J. B.; Ryu, C. K. Continuous
operation of the potassium-based dry sorbent CO2 capture process
with two fluidized-bed reactors. Int. J. Greenhouse Gas Control 2007, 1
(1), 31–36.
(14) Park, Y. C.; Jo, S. H.; Ryu, C. K.; Yi, C. K. Long-term operation
of carbon dioxide capture system from a real coal-fired flue gas using dry
regenerable potassium-based sorbents. Energy Procedia 2009, 1 (1),
1235–1239.
(15) Lee, S. C.; Kim, J. C. Dry potassium-based sorbents for CO2
capture. Catal. Surv. Asia 2007, 11 (4), 171–185.
(16) Seo, Y.; Jo, S. H.; Ryu, H. J.; Bae, D. H.; Ryu, C. K.; Yi, C. K.
Effect of water pretreatment on CO2 capture using a potassium-based
solid sorbent in a bubbling fluidized bed reactor. Korean J. Chem. Eng.
2007, 24 (3), 457–460.
(17) Hirano, S.; Shigemoto, N.; Yamada, S.; Hayashi, H. Cyclic fixed-
bed operations over K2CO3-on-carbon for the recovery of carbon dioxide
under moist conditions. Bull. Chem. Soc. Jpn. 1995, 68 (3), 1030–1035.
(18) Hayashi, H.; Taniuchi, J.; Furuyashiki, N.; Sugiyama, S.; Hirano,
S.; Shigemoto, N.; Nonaka, T. Efficient recovery of carbon dioxide from
flue gases of coal-fired power plants by cyclic fixed-bed operations over
K2CO3-on-carbon. Ind. Eng. Chem. Res. 1998, 37 (1), 185–191.
(19) Ficicilar, B.; Dogu, T. Breakthrough analysis for CO2 removal
by activated hydrotalcite and soda ash. Catal. Today 2006, 115 (1 4),
274–278.
(20) Park, S. W.; Sung, D. H.; Choi, B. S.; Oh, K. J.; Moon, K. H.
Sorption of carbon dioxide onto sodium carbonate. Sep. Sci. Technol.
2006, 41 (12), 2665–2684.
(21) Lee, S. C.; Chae, H. J.; Lee, S. J.; Choi, B. Y.; Yi, C. K.; Lee, J. B.;
Ryu, C. K.; Kim, J. C. Development of regenerable MgO-based sorbent
promoted with K2CO3 for CO2 capture at low temperarures. Environ.
Sci. Technol. 2008, 42 (8), 2736–2471.
(22) Lee, S. C.; Kwon, Y. M.; Ryu, C. Y.; Chae, H. J.; Ragupathy, D.;
Jung, S. Y.; Lee, J. B.; Ryu, C. K.; Kim, J. C. Development of new
alumina-modified sorbents for CO2 sorption and regeneration at
temperatures below 200 °C. Fuel 2011, 90 (4), 1465–1470.
(23) Lee, S. C.; Kwon, Y. M.; Park, Y. H.; Lee, W. S.; Park, J. J.; Ryu,
C. K.; Yi, C. K.; Kim, J. C. Structure effects of potassium-based TiO2
sorbents on the CO2 capture capacity. Top. Catal. 2010, 53 (7 10),
641–647.
(24) Lee, S. C.; Chae, H. J; Lee, S. J.; Park, Y. H.; Ryu, C. K.; Yi, C. K.;
Kim, J. C. Novel regenerable potassium-based dry sorbents for CO2
capture at low temperatures. J. Mol. Catal. B: Enzym. 2009, 56 (2 3),
179–184.
(25) Seo, Y.; Jo, S. H.; Ryu, C. K.; Yi, C. K. Effect of reaction
temperature on CO2 capture using potassium-based solid sorbent in
bubbling fluidized-bed reactor. J. Environ. Eng. 2009, 13 (6), 473–477.
(26) Zhao, C. W.; Chen, X. P.; Zhao, C. S.; Liu, Y. K. Carbonation
and hydration characteristics of dry potassium based sorbents for CO2
capture. Energy Fuels 2009, 23, 1766–1769.
dx.doi.org/10.1021/ef200499b |Energy Fuels 2011, 25, 3835–3842
Energy & Fuels ARTICLE

(27) Zhao, C. W.; Chen, X. P.; Zhao, C. S. Study on CO2 capture

using dry potassium-based sorbents through orthogonal test method.
Int. J. Greenhouse Gas Control 2010, 4 (4), 655–658.
(28) Zhao, C. W.; Chen, X. P.; Zhao, C. S. Effect of crystal structure
on CO2 capture characteristics of dry potassium-based sorbents. Chemo-
sphere 2009, 75 (10), 1401–1404.
(29) Li, L.; Wen, X.; Fu, X.; Wang, F.; Zhao, N.; Xiao, F. K.; Wei, W.;
Sun, Y. H. MgO/Al2O3 sorbent for CO2 capture. Energy Fuels 2010,
24, 5773–5780.
(30) Park, J. H.; Beum, H. T.; Kim, J. N.; Cho, S. H. Numerical
analysis on the power consumption of the PSA process for recovering
CO2 from flue gas. Ind. Eng. Chem. Res. 2002, 41 (16), 4122–4131.
(31) Khatri, R. A.; Chuang, S. S. C.; Soong, Y.; Gray, M. Thermal and
chemical stability of regenerable solid amine sorbent for CO2 capture.
Energy Fuels 2006, 20 (4), 1514–1520.
(32) Lee, S.; Filburn, T. P.; Gray, M.; Park, J. W.; Song, H. J.
Screening test of solid amine sorbents for CO2 capture. Ind. Eng. Chem.
Res. 2008, 47 (19), 7419–7423.
(33) Xu, X. C.; Song, C. S.; Miller, B. G.; Scaroni, A. W. Influence of
moisture on CO2 separation from gas mixture by a nanoporous
adsorbent based on polyethylenimine-modified molecular sieve
MCM-41. Ind. Eng. Chem. Res. 2005, 44 (21), 8113–8119.
(34) Han, G. Y.; Lee, G. S.; Kim, S. D. Hydrodynamic characteristics
of a circulating fluidized bed. Korean J. Chem. Eng. 1985, 2 (2), 141–147.
(35) Ruthven, D. M. Principles of Adsorption and Adsorption Processes;
John Wiley and Sons, Inc.: New York, 1984.
(36) Atkins, P. W. Physical Chemistry; Oxford University Press:
Oxford, U.K., 1990.
(37) Lee, S. C.; Choi, B. Y.; Ryu, C. K.; Ahn, Y. S.; Lee, T. J.; Kim, J. C.
The effect of water on the activation and the CO2 capture capacities of alkali
metal-based sorbents. Korean J. Chem. Eng. 2006, 23 (3), 374–379.
(38) Okunev, A. G.; Sharonov, V. E.; Aristov, Y. I.; Parmon, V. N.
Sorption of carbon dioxide from wet gases by K2CO3-in-porous matrix:
Influence of the matrix nature. React. Kinet. Catal. Lett. 2000, 71 (2),
355–362.
(39) Sharonov, V. E.; Tyshchishchin, E. A.; Moroz, E. M.; Okunev,
A. G.; Aristov, Y. I. Sorption of CO2 from humid oases on potassium
carbonate supported by porous matrix. Russ. J. Appl. Chem. 2001, 74 (3),
409–413.
(40) Okunev, A. G.; Sharonov, V. E.; Gubar, A. V.; Danilova, I. G.;
Paukshtis, E. A.; Moroz, E. M.; Kriger, T. A.; Malakhov, V. V.; Aristov, Y. I.
Sorption of carbon dioxide by the composite sorbent “potassium carbonate
in porous matrix”. Russ. Chem. Bull. Int. Ed. 2003, 52 (2), 359–363.
(41) Wang, X. P.; Yu, J. J.; Cheng, J.; Hao, Z. P.; Xu, Z, P. High
temperature adsorption of carbon dioxide on mixed oxides derived from
hydrotalcite-like compounds. Environ. Sci. Technol. 2008, 42 (2), 614–618.

3842 dx.doi.org/10.1021/ef200499b |Energy Fuels 2011, 25, 3835–3842